Ejection Fraction: Misunderstood and Over-rated (Changing the Paradigm in Categorizing Heart Failure)

The long-standing emphasis on ejection fraction (EF) is misguided. EF is erroneously assumed to be a measure of myocardial contractility. Of greater concern is the widespread classification of patients with heart failure (HF) based on whether EF is preserved (HFpEF) or reduced (HFrEF). In fact, EF does not provide any specific information on causation or underlying mechanisms. We believe that a revision or abandonment of this nomenclature is warranted and that categorization of patients with HF should more strongly emphasize underlying pathophysiology

What EF represents

EF is a characterization of ventricular ejection: the stroke volume (SV) expressed as a fraction of end-diastolic volume (EDV). Knowledge of EDV is essential in order to translate SV expressed as a percentage (EF) into absolute SV, a quantity of more physiologic and clinical significance. Hence, our admonition that one should never mention EF without, in the same breath, diastolic volume.

In the 1950’s physiologists used the instantaneous change in SV/EDV (i.e. the change in EF) as a measure of change in contractility under conditions of constant load. Over time we have largely forgotten that 1) EF is influenced by both preload (diastolic) and afterload (systolic) and cannot be interpreted as an index of contractility without knowledge of left ventricular (LV) loads and 2) structural changes leading to increases or decreases in LV EDV will strongly influence the EF at a given level of contractility and SV. So EF is twice removed from an index of contractility and has little meaning on its own.

Factors altering EF

Figure 1 portrays schematics of P-V curves selected to represent the uncoupling of EF from contractility. Panels A and B represent changes in preload and afterload, respectively, in a normal heart. Panels C and D represent cardiac remodeling in HF with cardiac hypertrophy, in C, and dilated cardiomyopathy in D. All panels include the same “control” P-V loop, in black, showing a 50 ml SV and 50% EF. In A, the P-V loop (in red) shows an increased EDV (preload), which increases SV and contractile force via the Frank-Starling mechanism; yet EF is barely changed (52%). In B, the increased vascular resistance (afterload) increases arterial pressure and contractile force (Anrep Effect), maintaining SV and increasing stroke work; yet EF is again barely changed (48%).

Figure 1.

Schematics of changes in P-V (Pressure-Volume) loops of the LV (left ventricle) and in SV (stroke volume) and EF (ejection fraction). The upper panels indicate the effects of increases in preload (A) and afterload (B) shown in red vs. control P-V curves in black. Data to support such changes were published in J. Physiol. 48: 465, 1914 by Patterson, Piper, and Starling. The lower panels indicate the effects of cardiac remodeling in heart failure with hypertrophic cardiomyopathy in (C) and dilated cardiomyopathy in (D) with reduced and increased EDVs (end-diastolic volumes) respectively.

C and D represent cardiac remodeling in HF with decreases in myocardial contractility and in both SV and EF. Although SV is equally reduced to 25 mls in both C and D and cardiac output may also be equally decreased, EF is significantly lower in dilated cardiomyopathy (19%) because of the larger EDV compared to hypertrophic cardiomyopathy (28%) which encroaches on the LV cavity in diastole and reduces EDV. In patients with hypertrophic cardiomyopathy, EF may increase through remodeling and reduced ventricular cavity volume, rather than increased SV or contractility.

In these examples increased preload and afterload increase contractile force; yet EF is essentially unchanged. Conversely in hypertrophic vs dilated cardiomyopathy the decreases in SV, cardiac output, and myocardial contractility may be similar; yet EF is lower in dilated cardiomyopathy through structural remodeling causing a larger EDV.

Ventricular volumes, EF, and cardiac remodeling

A structural change in LV volume is a dominant determinant of EF. The term “remodeling” arose to describe structural changes, first described in the 1970’s, occurring following a large myocardial infarction [1]. Scarring and thinning of infarcted myocardium and myocyte hypertrophy with interstitial fibrosis in the non-infarcted myocardium drive increases in EDV, with similar changes occurring in dilated cardiomyopathy [2]. The term remodeling may also apply to cardiac concentric hypertrophy and fibrosis without dilatation, often in response to a chronic increase in afterload, as seen in hypertension and aortic stenosis. Similar remodeling may also be due to a myocardial process, often associated with elements of the metabolic syndrome (diabetes, obesity, hyperlipidemia and hypertension) and advanced age [3].

Each of the above-mentioned conditions is generally characterized by reduced myocyte contractility, as well as reduced lusitropy. However, EF change may not reflect contractility as much as it reflects ventricular remodeling. (Strain measurements may provide more meaningful information regarding contractility.) Although in patients with reduced EF, inotropes may acutely improve symptoms, these agents may worsen the underlying disease process and increase mortality. In contrast, neurohumoral and vasoactive interventions that reverse remodeling also reduce mortality [4].

Inaccurate use of EF to distinguish systolic and diastolic dysfunction

Pathologic hypertrophy and fibrosis, occurring in most patients with HF, are associated with abnormal contractility and relaxation. It is a misconception that reduced EF equates with systolic dysfunction and preserved EF with diastolic dysfunction. Even more unfortunate has been primarily using EF to classify patients with HF in clinical trial design. As inotropes were explored to augment contractility in patients with HFrEF, calcium channel blockers were investigated to augment relaxation in patients with HFpEF. It was not recognized that the nature of remodeling – with or without LV dilation – was the primary driver toward reduced vs. preserved EF. Assuming that EF defines the underlying disease mechanism has resulted in numerous neutral and inconclusive clinical trials.

Should the metabolic syndrome be defined as a specific form of HFpEF?

The “metabolic syndrome” drives a process, perhaps accelerating one often seen in the elderly, leading to a form of HF with preserved EF with specific characteristics [5]. The pathologic process is one of tissue oxidation and inflammation, leading to myocardial fibrosis and hypertrophy as well as renal and vascular pathology [3]. This disease state, which we may call metabolic-senile cardiovascular disease (MetS-CVD), should not be labeled HFpEF but characterized specifically as HF with non-enlarged LVEDV, concentric myocardial hypertrophy and/or fibrosis, abnormal LV relaxation and/or compliance, left atrial enlargement, and elements of the metabolic syndrome and/or advanced age, in the absence of myocardial or pericardial infiltrative disease.

Conclusion

LVEF has exhausted its usefulness as a presumed marker of contractility and as a means of categorizing cardiomyopathies. In fact, the latter practice has stymied advances in pathophysiologic understanding and therapeutics. Underlying disease states may cross arbitrary EF boundaries, and multiple diseases may cause HF within any particular EF range. It’s time for a new paradigm. The descriptive term “HFpEF” is clinically and therapeutically useless. The term MetS-CVD distinguishes a circumscribed disease and pathological process, beginning before the clinical expression of HF, which should more aptly direct future therapeutic investigations. LV volume measurement is useful in assessing SV and in characterizing LV remodeling. But let’s focus on the underlying disease in considering novel therapies.